How we evaluate portable power stations
This page describes the method behind every comparison on the site. Individual articles state which parts of it applied to them.
How products are selected
We start from the job rather than the product. For each guide we define the loads a reader is realistically trying to power, then shortlist units whose inverter output and capacity can actually serve them. A unit only makes a guide if it is clearly the best answer for some group of readers. Products that would simply pad the list out to ten are left out.
What we compare
- Inverter output. Continuous watts, and whether any advertised surge mode applies to motor loads or only to resistive ones. This decides what a unit can start.
- Usable capacity. Rated watt-hours, with the caveat that usable capacity is always lower once inverter losses are counted.
- Recharge speed. How quickly a unit refills from mains, which decides whether it can use short windows between outages.
- Solar input ceiling. The cap on how fast a unit can accept solar energy, which matters more than battery size on trips longer than two nights.
- Weight. Reported as watt-hours per pound for guides where the unit gets carried.
- Switchover time. For backup use, whether equipment stays running through a cut.
How we calculate runtime
We do not reproduce manufacturer runtime charts, because those generally assume a constant draw. Real appliances cycle. We calculate from rated capacity against a realistic duty cycle, state the assumption in the article, and treat the result as an estimate rather than a measurement.
What our evaluation labels mean
- Hands-on testing. Someone on the team used the unit and the review names who, for how long, and in what conditions.
- Long-term ownership. A team member has lived with the unit over an extended period.
- Specification comparison. We did not use the unit. It was assessed by comparing published specifications against units we have used.
- Research based. We did not use the unit. The assessment draws on manufacturer documentation, owner reports and published expert reviews.
- Editorial selection. We did not use the unit. It is included because it covers a use case the rest of the shortlist leaves open.
What we will not do
We do not describe a unit as tested unless it was. We do not publish ratings or review counts we did not generate. We do not show a price we have not checked recently. Our publishing system enforces each of these automatically, and an article that breaks them cannot be built.